Heterogeneous crust and upper mantle across southern Kenya and the relationship to surface deformation as inferred from magnetotelluric imaging

Abstract

[1] We have used magnetotelluric data imaging to determine the resistivity structure across southern Kenya and our results suggest the presence of a buckled blocky or segmented lithosphere across the region. Prominent steep conductive zones at the Oloololo (OLO) escarpment and eastern rift margin allow us to subdivide the region into three crustal domains. West of OLO, a bow-shaped conductor underlies a 10 km thick resistive upper crustal unit spatially correlating with an exposed Archaean greenstone belt. Between OLO and the eastern rift margin are found steeply dipping alternating conductive and resistive zones that appear buckled. East of this belt are found prominent, 5 to 20 km deep, subhorizontal conductors atop steep resistive blocks with flanking conductors. The main steep features in the crust appear to extend below the seismic Moho and thus suggest the presence of anomalously thick crust across the region. A 50 km-wide and 4-8 km deep w-shaped (double half-graben) structure is suggested at the position of the Kenyan rift. We show that our inferred lateral zoning is consistent with collocated gravity and seismic measurements. We propose a link between the deep resistivity heterogeneity and surface deformation pattern in the area.

Making and Breaking of a Continent: Following the Scent of Geodynamic Imprints on the African Continent Using Electromagnetics

“…In the upper layers a rift-parallel conductivity anomaly occurs, which seems to be associated with the rift sediments. Although quite different in shape, Meju and Sakkas (2007) found an upper mantle high conductivity zone on the western side of the rift. Conductive blocks east of the rift valley were interpreted as the location of the suture zone between the craton and the Mozambique Mobile belt.…”

“…These pre-existing structures in the sheared Mozambique Mobile Belt have controlled the development of rifting and fracture propagation at the upper crustal scale. Meju and Sakkas (2007) re-analysed the KRISP-94 data set and proposed a 2D interpretation, after thorough strike and dimensionality analysis and subsequent down-weighting data with 3D effects. The 2D inversion model shows a very heterogeneous lower crust and upper mantle.…”

Heterogeneous crust and upper mantle across the SE Brazilian Highlands and the relationship to surface deformation as inferred from magnetotelluric imaging

“…Although static shift in dual‐polarisation data arising from near‐surface small‐size 3D bodies may be estimated by regularized inversion of linear regional profile data in ideal situations [ de Groot‐Hedlin , 1991], we opted for the alternative approach which uses local a priori information furnished by collocated transient electromagnetic (TEM) soundings to remove or reduce this uncertainty in MT data before interpretation [ Meju et al , 1999, 2003; Sakkas et al , 2002; Mohamed et al , 2002; Meju and Sakkas , 2007]. A follow‐up survey was therefore conducted in 2004 at the MT sites to acquire TEM data.…”

“…Here, we investigate the deep structure beneath the southeastern Brazilian Highlands and the spatial relationship to the outcropping linear geological features so as to shed some light on the above questions. Since this region is dominated by steep belts of strong ductile deformation [ Cunningham et al , 1996, 1998; Alkmim and Marshak , 1998], we have opted for a magnetotelluric (MT) approach in this study taking advantage of the technique's proven ability to detect strong lateral changes in subsurface structure and composition in steep basement terrains [e.g., Chen et al , 1996; Ledo and Jones , 2002; Unsworth et al , 2004, 2005; Bologna et al , 2005, 2006; Padilha et al , 2006; Meju and Sakkas , 2007]. First, we analyze MT data to reveal the resistivity structure along a 250 km transect (Figure 2).…”

Regularized extremal bounds analysis (REBA): An approach to quantifying uncertainty in nonlinear geophysical inverse problems

“…The other philosophy relies on linearization about a specific earth model and generates extreme models that satisfy a threshold data fit determined by the uncertainties in data constraints [ Meju , 1994; Kalscheuer and Pedersen , 2007]. In this particular approach, the emphasis so far has been on the influence of observational errors in nonlinear extremal inversion [e.g., Meju and Hutton , 1992; Meju and Sakkas , 2007], neglecting the other potential sources of uncertainty in subsurface parameter estimates such as model parameterization and predictive uncertainty of the overly simplified discrete models used to describe the underlying physical processes in complex geological media. It will be better to also take into account our modelling errors and a priori prejudices in the extremal inversion process.…”